My field data collection involves randomly placing quadrats throughout the Garry oak woodlands of Beacon Hill Park in Victoria, B.C. stratified to get equal representation of 3 different relative levels of shrub cover, then doing 10-minute point count surveys looking for presence/absence of spotted towhees (Pipilo maculatus) at the center of each quadrat within a 25 m radius. I’m not looking at the type of shrub cover or shrub species since towhees are ground foragers that nest low to the ground or on the ground in various types of vegetation. I’m repeating point count surveys at each station on 3 different days.
I’m sampling a total of 18 independent replicates and taking repeated data measures on 3 different days. There are 6 replicate sites per shrub cover category which are 5 m x 5 m quadrats gridded into 1 m x 1 m squares. The 3 categories are sparse (0-7 squares occupied out of 25), moderate (8-16 squares occupied out of 25), and dense (17-25 squares occupied out of 25). I’m aggregating the bird count data collected over 3 days for each of the 18 sites sampled so that for each site I can tabulate a proportion of days out of 3 where towhees were present during counts. This is so I can do a one-way ANOVA.
I’ve had some issues placing quadrats that are in higher foot traffic areas because I can’t leave them there for very long before it starts to get busy during the day. For these areas I place them and take them down with each visit which is more tedious and means I have to go back to those site coordinates exactly to ensure I don’t introduce another source of human error. What has helped is saving my location at these sites in google maps when I’m there so I can follow the directions back to it. I have the coordinates recorded in google maps for all 18 sites. Another issue I’ve had is with sighting towhees because they often like to hide in dense shrubs. If I hear only a towhee call and it’s obviously close-by to me (within 25 m, e.g. from in a nearby bush) but I can’t physically see it, then I still record it as present. This is one reason why I switched to only recording their presence/absence because counting how many exactly is unreliable when they could be moving around/not always visible.
My adjusted hypothesis is that the occurrence of spotted towhees (Pipilo maculatus) is influenced by the local density of shrub cover. So far, the patterns I’m seeing while doing my field data collections seem to support my prediction that spotted towhees will be present more frequently in the denser shrub cover sites. Most of my sightings have been at sites I classified as moderate or dense shrub cover, which is in keeping with their typical species behaviour as ground foraging birds that nest in shrubs/thickets.